Skip to main content

To make big data pay off, put decisions first

(Image credit: Image source: Shutterstock/wk1003mike)

Many organisations have a big problem with Big Data. Over 85 per cent of organisations tell Gartner that they are unable to exploit Big Data for competitive advantage, while at the same time 95 per cent of IT decision makers expect Big Data volumes and the number of data sources they use to grow further. The overwhelming majority of businesses are, therefore, already struggling to handle the amount of data that’s pouring in, and the problem is only going to get worse. 

What are businesses doing wrong? How can they avoid wasting their investments in storing and accessing data?

Refocus your goal: Decisions first

First, the idea that there’s a grand, one-size-fits-all solution – a ‘Big Data Holy Grail’ – is false. Rather than seeing Big Data as an end in itself, business leaders should focus on data-driven decision strategies. Rather than asking, “What can we do with all this data?” they should ask, “What decisions can we improve?” Then, get the data that could add insights to those decisions. This ‘decision-first’ approach emphasises not the data itself, but what enterprises want to achieve: the business end goal. 

Ultimately, you can’t pull value from Big Data until you know what decision you need to make, and why you need to make it. 

Make decisions a discipline

Second, organisations searching for the Big Data Holy Grail generally lack the analytics and processes needed to make the data work for them. Analytics separate the signal from the noise, and if you can’t do that you’ll make decisions based on inaccurate or incomplete information, which ultimately has a negative impact on customers. 

At FICO we have developed a six-step framework for best-practice decision-making that aims to create a full cycle with a feedback loop, so that every decision is better informed than the previous one and incremental improvements enhance the process every time. 

Here are those six steps: 

  • Codify the decision process and domain expertise so both can be easily examined, repeated and shared. FICO helped shape a recent industry standard for Decision Model Notation, which provides a common methodology for defining decisions.
  • Record the decision and the factors and data that led to it.
  • Model the insights with predictive analytics that can operate within your decision streams.
  • Optimise results through analytics that determine the best decisions to take given business goals and constraints.
  • Adapt models so they can be applied to multiple decision scenarios.
  • Improve decisions by simulating outcomes, measuring results, evaluating successes and optimising further.

Southwest Airlines, the world’s largest low-cost carrier, achieved real results for its customers and its business by doing just this.

Case study: Southwest Airlines

Today, Southwest is an airline that serves 100 million customers annually, but in its early days, it had just three aeroplanes serving Houston, Dallas and San Antonio. At that time, it was much easier for airline staff to manage crew assignments and ground operations manually in spreadsheets or on paper. 

As Southwest has expanded, the demands on its staff to manage an increasingly complex operation as smoothly and efficiently as possible have grown. Today, the airline has more than 600 planes that make over 3,500 flights every day to 93 destinations. This mammoth operation is handled by 45,000 employees – including 7,000 pilots and 11,000 flight attendants. 

In order to better manage its growing operations, Southwest implemented a 360° feedback approach to decision making. It deployed a solution, the Baker Airlines Operations Recovery Optimizer (known as the Baker), that utilises several algorithms, and hundreds of carefully calibrated parameters to improve real-time performance and quality, integrated decision-making. 

It’s the industry’s first real-time, integrated recovery solver that generates solutions to operational disruptions such as maintenance problems and weather events, while minimising the impact on passengers, flight and ground crews. 

Phil Beck, Manager, Optimisation Solutions at Southwest, explains: “Prior to the Baker implementation, superintendents of dispatch would address airline problems with a manual, labour-intensive process that could take hours to work out a single solution. Not only does this tool enable them to quickly react to problems within minutes, but they can also get ahead of potential disruptions hours in advance and have time to evaluate multiple scenarios. This project has supported the company’s goals and philosophies related to passenger hospitality, cost avoidance and operational performance.”

Since implementing the Baker, Southwest’s on-time performance is 10-15 per cent higher during times of irregular operation (e.g. uncontrollable changes such as weather conditions) compared to similar events in the past. Southwest has also improved customer satisfaction and made significant cost-savings in these adverse conditions by reducing flight diversions, tarmac delays and crew changes. In addition, performance during times of normal operations improved by more than 2 per cent in the last year alone. 

“The FICO Xpress Optimization Suite underlies some of the biggest, most expensive problems that we solve at the airline,” says Rusty Burlingame, Senior Operations Research Developer at Southwest Airlines. “There are so many opportunities for optimisation. We have been able to break into crew planning and flight planning, fuel purchasing, gate assignment, ground ops assignment and provisioning the aircraft.”

Whatever industry you work in, developing and implementing a reliable decision-making process can help drive business efficiencies. Southwest has put this into action in the transport sector, showing how important a reliable decision-making process is to take advantage of potential business benefits. It’s impossible to compare the paper and spreadsheet-based system the airline historically used to its state-of-the-art solution today. 

Decisions before data 

When Southwest was awarded the 2016 FICO Decisions Award for Decision Management Innovation, judge Jim Bander of Toyota Financial Services noted that “Southwest achieved an industry-first solution to a well-studied problem.” That’s what Big Data, when applied in a clear framework aimed at improving a defined decision or set of decisions, can achieve. 

The Big Data Holy Grail doesn’t exist, but there are still big benefits to be gained from using Big Data. The key is to sharpen the way you make decisions, and then Big Data will also help you to fly. 

Dr Stuart Wells is Executive Vice President, Chief Product and Technology Officer at FICO
Image source: Shutterstock/wk1003mike

Dr Stuart Wells is Executive Vice President, Chief Product and Technology Officer at FICO. Dr Wells previously held senior leadership roles with Sun Microsystems, Avaya and other top technology businesses.